CHAPTER X

LOGIC OF DISCOVERY VERSUS LOGIC OF JUSTIFICATION

Many of the debates in the preceding chapters are dependent

upon an inadequate understanding of ' discovery' and 'justification' .

The seemingly contradictory views expressed and the subsequent logical

consequences thereof are based on excessive exphasis either on the logic

of discovery or on the logic of justification. In what follows, an

attempt is made to clarify the two concepts and show how in the context

of the debate between realism and anti-realism, the dichotomy between

the two in unwarranted.

'Science progresses by correcting concepts' , says Mary Tiles.1

Although prima facie, the statement seems to be of common place,

historically viewed in the context of philosophy of science, it is radical.

Philosophy of science for the last two decades views 'progress' or

' change' as either shifts in research traditions or paradigms with no

cumulation of scientific knowledge. (One could call such a view 'majority'

position). Correcting concepts will involve neither historically descriptive

accounts offered by historians or sociologists of science (analysis in the

context of discovery) nor of the ahistorical, formally analytic kind

offered by logically oriented philosophers of science (analysis in the

context of justification). Mary Tiles, E. Pietruska-Madej2

and G.A. Van

Peursen3

are some of the philosophers who seem to be attempting to kill

the dichotomy in the context of scientific realism.

An inquiry into the nature of the 'logic of discovery' has to

begin with the historical development that gave rise to it. The question

concerning the 'aim of philosophy of science' is intrinsically linked up

2

with the ' justification versus discovery' debate, as it was traditionally

viewed that philosophy of science be concerned with the logic of justifi-

cation and it was the concern of psychologists, historians and sociolo-

gists to study the logic of discovery. Commonly understood, the logic

of justification concerns with rules of assessing and evaluating the

scientific claims, whereas the logic of discovery concerns itself with the

rules of generating new knowledge. The controversy in the present

context is not whether the 'logic of justification' is the legitimate aims

of science. The question is whether that is the only aim of philosophy

of science. Further, whether 'discovery' is ever the concern of

philosophers and whether there is logic in discovery, have become the

central issues in the present debate. And the position one adopts

regarding the above issues will throw light on the question regarding

scientific realism. In short, are there two seemingly opposed 'logics' ,

one for discovery and the other for justification .

The debate regarding the primacy of logic of justification over

logic of discovery (even the question whether there is a logic of

discovery) is as old as philosophy of science. An overview of the

debate at its best is provided by Martin V. Curd4

' who observes three

approaches to the problem. The 'logical positivists' like K.R. Popper 5,

C.G. Hempel6

and R. Braithwaite7

who accept hypothetico-deductive account

of science and do not accept any possibility of logic of discovery . H.

Reichenbach8

and W. Salmon9

also do not accept logic of discovery, but

unlike logico-empiricists, they accept an inductive-probability account

10 of science. C.S. Pierce, N.R. Hanson

11 and other Kuhnians who

accept abductive inference account of science, argue for greater emphasis

on the logic of discovery which is as important (if not more) to account

2" 9 for change and ' progress ' in scientific theory .

Karl Popper , C .G . Hempel and Hans Reichenbach had strong

reservations regarding the logic of discovery . Popper ' s rejection of the

ty of rmy logic of cl I...cove-17y is a natural consequence of the fact

that for him the work of scientists is to put forward theories and testing

them . "The act of conceiving or inventing a theory" , according to

Popper "is ... neither to call for logical analysis not to be susceptible

of it . The question how it happens that a new idea occurs to a man -

whether it is a musical theme, a dramatic conflict , or a scientific theory -

may be of great interest to empirical psychology ; but it is irrelevant to

the logical analysis of scientific knowledge there is no such thing

as a logical method of having new ideas , or a logical reconstruction of

this process . (Popper' s) view may be expressed by saying that every

discovery contains ' an irrational element ' , or ' a creative intuition ' in

Bergson' s sense" .12

C .G . Hempel and R . Braithwaite accept Popper ' s

position namely , the role of science is to explain the phenomenon and

consequently the concern is with the logic of justification .

This hypothetico-deductivist position, particularly that of sKarl

Popper has been labelled • as ' nihilistic attitude ' towards scientific

13 . discovery by some philosophers . E . Pietruska-Madej points out that

Popperians do not see any philosophical problems in scientific discovery .

Philosophy of science which employs theoretical tools of logic cannot say

anything about the process of theory-generation which is an empirical

issue according to deductivists

If Popper is called hypothetico-deductivist , Hans Reichanbach is

a probability-inductivist as his position is based upon acceptance of an

inductive-generalization model for the generation of new knowldege in

science. Hans Reichanbach first used the terms 'context of discovery'

and the 'context of justification' with a view of distinguishing between

psychological description of thought processes as a temporal sequence of

steps and logical relations between propositions. Reichanbach expressed

the impossibility of a logic of discovery when he says:

"The mystical interpretation of the hypothetical-

deductive method ci.s an irrational guessing springs

from a confusion of context of discovery and the

context of justification. The act of discovery

escapes logical analysis; there are no logical rules

in terms of which a "discovery machine" could be

constructed that would take over the creative function

of' the genius. But it is not the logician's task to

account for scientific discoveries; all that he can do

is to analyse the relation between given facts and a

theory presented to him with the claim that it explains

these facts. In other words, logic is concerned with

the context of justi.fication. And the justification of a

theory in terms of observational data is the subject of

the theory of induction".14

Reichanbach's rejection of 'logic of discovery' is based on his

understanding of 'logic' which for him is a "normative enterprise

concerned not with the description of inferences that people actually make

but with their justification through the provision of a rational reconstruct-

15 ion . The "manner in which the logical inferences are actually made

is strange and obscure and rarely resembles the formal method of logic" 16

points out Reichenbach. It is therefore by definition impossible for

Reichanbach to have a logic of discovery as there can be no logic of the

context of discovery. This does not mean that for Reichanbach all

discoveries are consciously made inductive inferences - what he necessarily

0 ", 4, I

admits is that scientific discoveries (like that of Einstein) could be

rationally reconstructed within the context of justification.

The hypothetic-deductivists (like Ilempel and Popper) and the

probability inductivists (like Reichenbach) do not reject the role of

'discovery' in scientific theorizing. What they seem to assert on the one

hand that the concern of philosophy of science is the logic of justifi- r

cation and consequently any discussion regarding the context of discovery

should be subsumed under the logic of justification.

N.R. Hanson alleges that the hypothetic-deductivists have ignored

discovery by calling it psychological, historical or sociological.

Hanson believes that hypotheses, theories, laws and observations have

been created, innovated and developed by minds that are endowed with

creativity . "Physicists", he says, "do not start from hypothesis; they

start from data. By the time a law has been fixed into an H-D system,

really orginal thinking is over. The pedestrian process of deducing

observation statements from hypotheses comes only after the physicist

sees that the hypothesis will at least explain the initial data requiring

explanation. This H-D account is helpful only when discussing the

argument of a finished research report, or for understanding how the

experimentalist or the engineer develops the theoretical physicist' s

hypotheses; the analysis leaves undiscussed the reasoning which often

points to the first tentative proposals of laws ... The initial tuggestion

of an hypothesis is very often a reasonable affair. It is not so often

affected by intuition, insight, hunches or other imponderables as bio-

graphers or scientists suggest. Disciples of H-D account often dismiss

the dawning of an hypothesis as being of psychological interest only, or

else claim it to be the province solely of genius and not of logic. They

are wrong. If establishing an hypothesis through its prediction has a

logic, so has the conceiving of an hypothesis".17

Hanson proposes 'retroactive inference' by means of which he can

account for creative scientific discoveries. 'Retroactive inference' is

said to be neit her inductive nor deductive inference. C.S. Pierce

proposes abductive inference' (Hanson claims that his concept is based

on Pierce's) which has a similar claim - that the logic employed is

neither inductive nor deductive. Pierce's 'abduction' (which has many

interpretations) has the following logical form:

"The surprising fact, C, is observed.

But if A were true, C would be a matter of course.

Hence, there is reason to suspect that A is true". •

Pierce's labductive inference' suffers from many lacuno.e,which

critiques pointed out lead to ' psychologism' . Although Hanson tries to

overcome these lacunae, his 'retroactive inference' has been criticised

on the ground that he does not account for the background that gives

rise to inferences.19

Further, Hanson fails to distinguish between logic

of generation and the logic of evaluation. And it is in this context

that Curd's argument regarding the logic of generation deserves

attention.

Martin V. Curd's analysis of three approaches is based upon

certain distinctions he has made. These distinctions are later on

employed to review the threefold aspects regarding the possibility of

'logic of discovery' . It may be noted at this stage that the debate

regarding the 'logic of discovery' versus 'logic of justification' is not

concerned with things but theories. And for 'Curd, a discovery of a

18

fr'; f) 0

scientific theory does not refer to a specific moment when a theory is

discovered, but to the period of theory generation . Curd defines this

period as the time "beginning at the moment when a scientist (or research

group) first begins thinking seriously about a problem and ending when

the theory ... is first written down in a form suitable, say for publication

in a scientific journal" .20

Curd further observes two different aspects of

the period of theory generation; namely, one, the logic of prior assessment

and two, the logic of theory generation . Curd observes two kinds of

logic of prior assessment (i.e. "methodological appraisal of hypotheses

after they have been generated but before they have been tested") :

(a) the logic of probability; and (b) the logic of pursuit . The logic

of probability concerns itself with the questions of which hypothesis is

likely to be true; whereas the logic of pursuit concerns itself with the

questions of which hypothesis should a scientist begin to work on. In

other words a scientist may begin with a hypothesi'S for reasons other

than likely truth of the hypothesis . Curd' s interest is in the logic of

pursuit which he claims has more immediate practical relevance to

scientific inquiry than the logic of probability. It is, according to

Curd, only "workable notion of a logic of discovery in the sense of a

logic of prior assessment that one can formulate".21

Unlike our

traditional conception where the 'logic of pursuit' would be parasitic

upon the 'logic of probability' , Curd claims that rarely can we talk

about our' hypothesis as likely to be true or probable, we however,

"have the means for deciding which hypothesis to pursue, which

decisions do not depend on our inability to make these probability

judgements" .22

Curd after the analysis of the three approaches concludes that

f)t--,!•-) .

the objections raised do not establish the impossibility of the logic of

discovery as either a logic of. prior appraisal or as a rational recon-

struction of inferences to the theory. Further, the only "sensible logic

of prior appraisal is the logic of pursuit which provides the justification

for inferences to the theory in our rational reconstruction of the period

of theory generation ... The factors that justify our inferences to the

theories in the first place are the same as those that were used to

decide which theory to pursue after they have been generated".23

As mentioned in the beginning of this chapter, G.A. van Peuersen

attempts to kill the traditional distinction between the context of

discovery and the context of justification. Peuersen discusses three

aspects of the scientific framework, which compels him to say that the

whole of scientific procedure is nothing but discovery. Consequently,

the dualism of contexts is deemed to be untenable.

The argument in brief consists in showing that:

1) The success or failure of natural and social sciences ita modern

technological society depends upon the social organizations, the financial,

and governmental policies, etc. It is this that compelled Popper and

others to recognize 'the irrational element' in the process of discovery.

This is the diachronic conception of discovery as against synchronic

context structure of justification. Justification, in synchronic context,

becomes "a construction in retrospect, functioning only as a stabilizing

factor, inhibiting renewal and advancement".24

This shows the

estrangement between traditional methodology and heuristic practice.

2) Secondly, Peuersen argues that there is a feeling of 'loss

of reality' of the daily course of events confronted by the scientifically

controlled character of the modern society. Such a feeling is fostered by

the scientific method leading to "scientific explanation and concrete

reality (which) is an ontological issue since the reference of a scientific

discipline to reality is at stake".25

3) Thirdly, Peuersen notices a number of cultural constraints in

the framework of contemporary scientific enterprise. T.S. Kuhn, A.C.

0.-ed Crombie and others mentioned as philosophers representing this third

aspect of the framework of science. Peuersen believes that "the real

issue at stake is the tension between the universal claim of scientific

procedure and the specific, often hidden, convictions and world-views of

a certain culture. The investigation of such ,a cultural context might

stimulate the search for new dimensions of theory and concept formation.

It can lead, in this way to new heuristic devices.26

The study of Peuersen's three aspects of framework of scientific

enterprise reveals: (a) dualism between two contexts (discovery and

justification) is untenable; (b) because of the conflict between the

traditional methodology and heuristic practice, there is a need of

integrating the two contexts; and (c) justification and discovery are

neither separated nor are merged into one another. Instead, "justification

as a sedimentation of discovery (on the one hand be treated as) the

outcome of the process of discovery (and on the other as) an adjusting

device within the ongoing discovery".27

The result of this analysis is

not blending of the two contexts, or absorbtion of one by the other.

What Peuerson has attempted is to create two contexts - but 'one wider,

less defined and demarcated from the wider field of daily culture', -

and the other 'more restricted, functioning as a kind of safety device on

behalf of the wider strategy of discovery' .

o

Mary Tile's claim lhol sciencr progre,.-,se,, by correcting W.-,

concepts and the framework offered by her allows for a cumulative model

of scientific progress against Kuhnian 'incommensurable alternatives' .

Mary Tiles observes three levels or dimensions to a scientific concept:

the empirical, the theoretical and the metaphysical. The empirical

dimension (often taken for granted) concerns itself with the concepts

that are empirically developed in a scientific context. But since the

object of scientific investigation is "never given purely empirically",

the concept thereof has to accommodate the changes occurring in both the

theoretical and the metaphysical dimensions of science. This accounts

for the non-empitical framework determining what kind of a thing or

phenomena it is that is investigated.

Although one can never claim that a scientific concept is solely

determined by its theoretical dimension, the theoretical component of

scientific knowledge, it is epistemologically important to account for how

a new theory forms an inseparable part and is related to the existing

empirical framework (knowledge). Mary Tiles, claims that "theories are

attempts to fill in the gap between emperical and metaphysical dimensions

of the concept, a gap which is already structured by its end points".28

At the empirical level the correcting of concepts occurs due to the demand

for increased precision, at theoretical level the correcting mechanism acts

due to demand for increased logical precision and conceptual clarity.

Mary Tiles observes two kinds of 'inbuilt mechanisms' for conceptual

change: one, "the experimentally based sequence of empirical classification

associated with increased accuracy in measurement"; and two, "the

sequence of ever more abstract and general theoretical concepts".29

These

two are not independent sequences according to Tiles. The third

r) 0

dimension, namely metaphysical dimension of her analysis is questionable

within the context of science. The metaphysical dimension of scientific

concept is observed from the fact that theories are scrutinized in the

light of explanatory ideals which are based on metaphysical views

regarding structure and nature of reality. Metaphysical views which on

the one hand determine or act as constraints on the empirical and

theoretical dimension of the concept, have to ensure mechanisms of self-

correction on the basis of new theoretical or empirical (experimental

evidence). Each of the three levels (empirical, theoretical and meta-

physical) have, specific forms of concept correction. But, as Mary Tiles

points out, "the levels are not mutually independent in that both the

stimulus and the means for correction on one level are provided by work

on another. It is in the direction of increased coordination and

unification of the initially very disparate levels that science progresses". 30

One point remains unclear, namely, what in the ultimate constitutes

'metaphysical level' for Tiles? The suspicion that 'metaphysical' may

constitute more than nature and status of reality, is suggested by the

fact that she admits "that science is neither the original nor the only

source of world views".31

In a scenario full of constrasting and confusing frameworks whose

actual representations fall far short of experience of working scientists,

E. Pietruska-Madej's proposal provides both direction and insidht into

the plausible 'exit' . Pietruska-Madej was trying to get out of the

'disharmony' between two objectives of science. And she puts it: "the

word 'disharmony.' is justified because there is no way to cope success-

fully with the task of rationalising the historical process of the growth

of science, so long as one holds to the opinion that philosophy of science

2 &;

can and s-hould study only established and articulated knowledge. On the

contrary, it seems impossible to explain the growth of scientific knowledge

without taking into consideration its formation. It is impossible because

the second is a vital part of the first".32

Pietruska-Madej recognizes

the anti-psychologism attitude as the root-cause of nihilistic attitude.

Pietruska-Madej begins with the assumption that anti-pschologism in

epistemology should not exclude us from the entire process of the

emergence of new knowledge. Besides, there is much more to discovery

and generation of scientific knowledge than mer,ely psychological experience

of the individual scientists. Scientific discovery is a very complex

phenomenon, having both subjective (what occurs in individual's mind)

and objective (typical characteristics of the background situation -

discovery generating situation) aspects.33

Taking clues34

from two celebrated scientists - scholars, P.D.

Medawar (Nobel Prize Winner for Medicine in 1960) and Derek J. de

Solla Price (historian of science), Pietruska-Madej takes as her starting

point the unexplained facts of 'logical necessity' of a scientific discovery

and 'lying waiting' to be discovered; and attempts to explain many a

phenomena igonred by philosophers of Science, such as "why does science

sometimes manage to ignore insights of genius only to rediscover them

years later? Why does it at other times explode with a series of

simultaneous discoveries? How should we account for 'premature'

discoveries which fail, how can we distinguish between discovery in the

subjective psychological sense and scientific discovery proper?',35

Pietruska-Madej believes that "there is some internal logical

determination in science by which science proceeds from established

knoNiledge to new elements of knowledge, and tha. t there comet,into play

0 c,)

something like ' necessity' conducive to a new discovery". 36

Secondly, science matures "for a given discov'ery, and that so long as

it is not mature enough for this (discovery) even the most ingenious idea

is doomed to be regarded as irrelevant for scientific progress". 37 And

finally, when a particular stage of science is reached, a discovery is

inevitable, such that even "several scientists may make the discovery

simultaneously". 38 These three concepts (logical necessity, maturity and

inevitability) seem to suggest that "there are certain objective intrinsic

conditions and logical mechanisms in science which determine the actual

substance of a newly discovered idea. Accordingly, providing accurate

definitions of these concepts seems to be a most urgent task, and one

which should be undertaken rather by philosophers of science than

psychologists".39

It is on the basis of the above analysis that

Pietruska-Madej argues that discovery is not a 'creative subjective act'

occuring in a vacuum. It occurs in an objective situation which both

influences and determines the emergence of a new idea. This analysis

further explains why some 'discoveries' or 'new ideas' fail to catch the

imagination of scientific community of a particular period. These

'discoveries' have 'to wait' for a more appropriate period (when the

science is more mature) to be accepted as discoveries.

Again Pietruska-Madej's analysis allows us to distinguish between

genuine scientific discovery and subjective novelty. Someone can discover

something 'new' for his own subjective consciousness which may not be

new to science' . To say an idea is new in scientific sense of discovery

depends upon many objective factors - "it is the objective situation in

science that determines the necessity, and the possibility, of the

emergence of a definite idea in science". 40

0 f:.; • r.,

The proces5 5cientitic discovery has non-psychological aspects

and hence cannot be construed as a psychological act or a flash of the

intuition of a genius. It therefore calls for an analysis not of the

psychological preconditions of the discoverer bUt for the reconstruction

of objective situation in science at a particular historical moment,

"such a situation"„ Pietruska-Madej points out, "is always characterized

by a set of accepted propositions which constitute science, ... by

logical relationships and mutual dependencies between the set's elements,

by a accepted epistemological values, by normative ideals for research

activity, and so forth".41

It is, therefore, the objective discovery -

generating situation that determines when and how a new idea appears

in science. Such a situation generates, determines the content and lays

down range of constraints on the new idea. lt is precisely because of

this that to understand a scientific discovery, we have to reconstruct

the discovery-generating situation the objective reasons why the new

idea is acceptable; the invetablility and necessity of the new ideas;

the internal and external situation of the science; the logical relations

between the various constituent elements of the situation (the laws,

empirical data, regulative ideals of scientific activity, etc.); the

logical corrolaries of the 'n,ew idea' etc.

It is admittedly true that for such a reconstruction of a discovery-

generating situation one must have recourse of historical materials. To

reveal the most general features of such situation and logically

reconstruct tha process of transition from 'old to new' scientific knowledge

is a task of philosophers of science. One does not inquire into a

mysterious 'discovery machine' or Bergson's 'creative intuition' while

inquiring into the logic of discovery. Studying the objective aspects of

0 ri (Zo (,)

the scientific discovery is one ofobjectives of philosophy of science . And

if one wants to inquire into the ' progress or change' in science one has

to inquire into both the logic of justification and the logic of discovery-

generating situation .

NOTES

1. Mary Tiles , "Correcting Concepts" , Ratio, 27,1,1985, pp . 19-36 .

2. E. Pietruska-Madej , "Should Philosophers of Science Consider

Scientific Discovery?" , Ratio, 27 , 1, 1985, pp . 7-18 .

3 •

G . A . van Peuersen , "Discovery as the Context of Any Scientific

Justification" , Man and World , 22, 4; 1989, pp . 471-484 .

4 .

Martin Curd, "The Logic of Discovery : An Analysis of Three

Approaches" , in Readings in the Philos'ophy of Science, ed. B. A.

Brody and R .E. Grandy, (New Jersey: Prentice-Hall, 1989),

pp . 417-430 .

5. K .R . Popper , Logic of Discovery , (London : Hutchinson & Co . ,

1965) .

6 . C .G . Hempel, Aspects of Scientific Explanation , (London : Collier- MacMillan , 1965 ) .

7 . R . Braithwaite , Scientific Explanation , (Cambridge : Cambridge

University Press , 1955 ) .

8 . Hans Reichanbach , Experience and Prediction , (Chicago : Chicago

University Press , 1938) .

9 . W . Salmon , "Bayes ' Theorem and the History of Science" ,

Minnesota Studies in Philosophy of Science, Vol . V , ed . R .

Stewer , (Minneapolis : University of Minnesota Press , 1970) ,

pp . 87-111.

10 . C .5 . Pierce, Collegeted Papers , eds . ,C . Hartshorne et al. ,

(Cambridge : Harward University Press , 19 )

2 3 5 11. N.H. Hanson, Patterns of Discovery, (Cambridge: Cambridge

University Press, 1958).

12. K.R. Popper, Logic of Discovery, p. 32.

13. E. Pietruska-Madej, "Should Philosopher' s of Science Consider Scientific Discovery?", p . 8-9.

14. Hans Reichanbach, "THe Rise of Scientific Philosophy", p. 231.

15. Martin Curd, "The Logic o Discovery : An .Analysis of Three Approaches", p. 423.

16. Hans Reichanbach, The philosophy of space and Time, (New York:

Dower & Co., 1957), p. 43.

17. N.R. Hanson, Patterns of Discovery, p. 70-71.

18. Quoted in Martin Curd, "The Logic of Discovery : An Analysis of three

Approaches", p. 418.

19. P. Achinstein, 'Law and explanation' (Oxford : Clarendan Press,

1970).

20. Martin Curd, "The LogiC of Dit-covery : An Analysis . of .Three Approaches", P . 418.

21. Ibid. p. 419.

22. Ibid,

23. Ibid, p. 427.

24. G.A. van Peursen, "Discovery as the Context of Any Scientific Justification", p. 47S.

2S. Ibid. p. 474.

26. Ibid.

27. Ibid. p. 482.

28. Mary Tiles, "Correcting Concepts", p. 27.

29. Ibid. p. 29.

30. Ibid. p. 20.

31. Ibid. p. 31.

32. E. Pietruska-Madej, "Should Philosopher's of Science Consider Scientific Discovery", p. 8.

33. Ibid. p. 12.

34. P.B. Medawar writes in The Hopes of Progress, that if J.D.

Watson and F. Crick did not dAover 'double helix of DNA' A

someone would have discovered it very soon - and discovered

only that and nothing else. Simdlarly Derek J. de Solla Price in

"what is the Difference between Science and Technics?" said

'had not Boyle not discovered his law, someone else would.

Every fact and every theory waits (depending upon the maturity

of science) to be discovered.

35. Ibid. p. 14.

36. Ibid. p. 15.

37. Ibid.,

38. Ibid.

39. Ibid. fp. 15.

40. Ibid. p. 16.

41. Ibid. p. 17.

f:tf, 0